Professional Documents
Culture Documents
3 BD 1 FC 8 A 79892 Ec 04 Abc
3 BD 1 FC 8 A 79892 Ec 04 Abc
Introduction
Libet-style experiments have received much attention in both neuro-
science and philosophy. In general, these experiments are taken to
show that if subjects are asked to freely decide when to move or what
movement to make, the conscious intention or urge to move and the
actual movement can be predicted on the basis of preceding uncon-
scious brain activity (e.g. Haggard and Eimer, 1999; Libet, 1985;
Correspondence:
Lieke Asma, VU Amsterdam, The Netherlands. Email: l.j.f.asma@vu.nl
not ‘I’.
To assess whether these findings really pose a threat to free will, it
is important to critically examine the definitions used in these experi-
ments. Neuroscientists typically define and operationalize voluntari-
ness first, and on the basis of the experimental results conclude
whether the voluntary action or decision was free as well.1 This is also
the case for Filevich, Kühn and Haggard’s (2013) experiment: sub-
jects have to voluntarily decide to veto or not and, on the basis of the
occurrence of unconscious neural precursors, it is concluded whether
the decision was also free. Both these definitions are identical in all
Libet-style experiments and have been criticized before (e.g. Mele,
2009; Schlosser, 2014). However, by analysing the free won’t experi-
ment I will show that the conditions for voluntariness and freedom are
more stringent than these philosophers take them to be. I will argue
that because the voluntary decision is (implicitly) operationalized as
not caused by external and internal factors and freedom is understood
as conscious initiation, the experiments are set up in such a way that
the conclusion that the brain ‘decides’ is almost unavoidable. This is
because the only decisions subjects are allowed to make in these
1 This distinction is for example made in Haggard (2008). Voluntary action and deciding
(also referred to as ‘intentional’ or ‘self-generated’, or ‘intentional inhibition’ in the
case of the veto decision) are defined in the neuroscientific literature without referring
to the causal role of consciousness, but by pointing out the role of internal and external
causes and the experience of control (e.g. ibid., p. 934). Whether these actions and
decisions are free depends on the role of conscious processes (e.g. ibid., p. 941, or
Haggard and Libet, 2001, p. 47).
10 L. ASMA
2 In fact in Filevich et al.’s (2013) study subjects did not have to refrain from acting
altogether, but had to decide to either act immediately or to delay their response — for
now I will take this to be a voluntary veto decision and I will not take into account the
question of whether this is a genuine veto decision.
THERE IS NO FREE WON’T 11
that case it would not be a veto decision. Subjects first had to acquire
the urge or intention to move and actually prepare to make the move-
Copyright (c) Imprint Academic 2017
3 Mele (2008) analysed Brass and Haggard’s (2007) study and concluded that it is not
clear what strategies participants are actually using in the experiments and, because of
that, whether the participants are really deciding not to act on an intention they
acquired. For the purposes of this paper, I assume that the experiment actually targets
veto decisions and that subjects are able to do and are in fact doing what is asked of
them. My goal is to critically examine the assumptions that play a central role in the
experiment and in the conclusion that is drawn.
4 A further problem that should be pointed out here is that the experiment, as well as the
regular Libet-style experiments, only establishes temporal precedence and the findings
do not allow for claims implying causal determination (see for example Fischborn,
12 L. ASMA
Voluntary Deciding:
Internally Generated or Underdetermined?
In this section, I focus on the definition of voluntariness that is used in
experiments on the veto decision. In his 2014 paper, Schlosser
addresses this definition in which a, according to him problematic,
distinction is made between internal and external factors that influence
the action and decision. Even though this is the way neuroscientists
explicitly define voluntariness, I show in this section that deciding
For personal use only -- not for reproduction
ble factors that may play a role from the experimental setting. Because
of this, making a voluntary veto decision amounts to subjects deciding
without being allowed to take any factor or reason into account.
In neuroscientific experiments on the veto decision, Libet’s original
definition of voluntary action is taken as a starting point. As he states:
(a) [the voluntary action] arises endogenously, not in direct response to
an external stimulus or cue; (b) there are no externally imposed
restrictions or compulsions that directly or immediately control sub-
jects’ initiation and performance of the act; and (c) most important,
subjects feel introspectively that they are performing the act on their
own initiative and that they are free to start or not to start the act as they
wish. (Libet, 1985, pp. 529–30)
Also in these more recent studies on the veto, internal generation is
central: ‘the process or signal that cancels or inhibits the action is not
the result of any external signal or instruction, but is generated
internally by the subjects themselves’ (Filevich, Kühn and Haggard,
2012, p. 1108). Furthermore, neuroscientists in the field link the veto
decision to self-control, i.e. the ability to postpone the immediate
gratification of a desire to act to achieve other benefits, which, in the
case of a voluntary decision, should not be the result of a response to
an external cue but generated internally (Kühn, Haggard and Brass,
2009, pp. 2834–5).5
2016). Since in the current paper my aim is to account for the occurrence of this
temporal precedence, I will not take this problem into account.
5 Examples that are used include refraining from sending an angry email (Filevich, Kühn
and Haggard, 2012), overcoming cigarette addiction (Kühn, Haggard and Brass, 2009),
THERE IS NO FREE WON’T 13
from reaching for a cup of coffee, because (a) I feel I do not want
more coffee, (b) I decided to drink less coffee a couple of days ago, or
(c) refraining from reaching for the cup was a spontaneous decision.
According to the explicit definition held by neuroscientists, my
decision to veto is voluntary in all these cases: the decision is not the
direct result of an external cue, it is generated internally, and I experi-
ence control over my decision — I feel that I could have reached for it
anyhow had I decided to do so.
However, in the experiments the subjects had to decide to veto their
initial urge or intention or not in circumstances that were much more
restricted than these descriptions allow. The subjects had to wait for
an intention or urge to move to emerge and then decide spontaneously,
without pre-planning, to veto or not. No external factors were present
or changed to make sure that they could not causally contribute to the
decision that was made.6 Furthermore, and this is where the
discrepancy between the explicit definition and the implicit operation-
alization becomes clear, the experiments are conducted in such a way
that no internal cues or factors were manipulated, measured, or
allowed to play a role either. There was no goal to be reached or
acting in socially desirable ways (Filevich, Kühn and Haggard, 2013), or resisting
criminal activities (Parkinson and Haggard, 2014).
6 Kühn, Haggard and Brass (2009) did integrate external cues in their experiment,
because previous studies on the veto ‘lacked the context of reasons for performing and
inhibiting action that generally exist outside the laboratory’ (ibid., p. 2835). However,
the subjects were not allowed to decide on the basis of these reasons whether to veto or
not. Later in the paper it will become clear why this is important.
14 L. ASMA
7 Schüür and Haggard (2011) also criticize this description. I use the remarks from their
paper only to clarify what underdetermined decisions are, and what the rationale is
behind operationalizing voluntariness this way.
THERE IS NO FREE WON’T 15
intention to move merely ‘bubbles up’ from our brain (Libet, 1999, p.
Copyright (c) Imprint Academic 2017
separate from the brain, for which no predictive neural correlates may
be found. Filevich et al. capture this intuition by stating that their
‘results suggest that an important aspect of “free” decisions to inhibit
can be explained without recourse to an endogenous, “uncaused” pro-
cess: the cause of our “free decisions” may at least in part, be simply
the background stochastic fluctuations of cortical excitability’ (ibid.,
p. 10).
Thus, the veto decision would only be free, according to neuro-
scientists’ definition of freedom, if the decision is consciously
initiated, which means that it cannot be predicted on the basis of pre-
ceding unconscious neural processes. That means that arguing that
conscious processes play a causal role, as Mele (2009) does, does not
fully target the stronger assumption that underlies the conclusion that
is drawn from the experiment.
our intuitive sense of free will’ (p. 194). He argues that our intuitive,
common-sense conception of free will involves a central role for con-
Copyright (c) Imprint Academic 2017
inclination to veto or not) and (b) the decision to veto or not should be
consciously initiated, or, in other words, not be predictable on the
basis of unconscious brain activity. It is important to realize that this
setting almost unavoidably leads to the conclusion that the brain
‘decides’ and that conscious deliberation is bypassed. I will provide
three arguments for this claim.
First, because the subjects had to make voluntary veto decisions of
the underdetermined kind, there was nothing to consciously deliberate
about in the experimental setting. The experiment was set up in such a
way that subjects did prepare to move and probably did acquire a
conscious urge or intention to move.8 Since the goal was to exclude all
external and internal cues from the experimental setting, that urge to
For personal use only -- not for reproduction
move is the only factor that may play a role in the decision to veto or
not. This means that there is simply nothing to deliberate about and
Copyright (c) Imprint Academic 2017
8 As said, Mele (2008) argued that this might not be what actually happens, see footnote
3.
9 This point, often put as Libet-style experiments being about the liberty of indifference,
has been made before in the literature, see for example Mele (2009, p. 84). However,
my aim is not mainly to criticize this definition of freedom directly, but to show that
thinkers who take the findings of Libet-style experiments seriously demand something
that will not occur in this experimental set-up.
THERE IS NO FREE WON’T 19
tion. The central idea in both definitions is that voluntary actions and
decisions should not be caused directly by external cues or at least
‘not uniquely specified by external cues’ (Schüür and Haggard, 2011,
p. 1698). In this section I examine these proposals and point out some
limitations. Furthermore, I elaborate on the fact that the definition of
freedom seems to be generally accepted.
In the case of operant action, distinguished from respondent action,
the criterion is that the subject is not directly responding to external
cues. Passingham and Lau (2006) describe free choice as ‘choices that
are not imposed on the subject by an external agent’ (p. 53). In a
similar vein, Filevich, Kühn and Haggard (2012, p. 1108) describe
operant veto decisions as decisions that are internally generated: ‘[b]y
definition, the process or signal that cancels or inhibits the action is
not the result of any external signal or instruction, but is generated
internally by the subject themselves.’ There are different ways to
interpret these descriptions, but the idea is that when internal factors
bring about the actions, they depend ‘on wide integration of informa-
tion beyond the current stimulus, such as memory of previous experi-
ence and predictions of future outcomes’ (ibid., p. 1116). A different
proposal comes from Schüür and Haggard (2011). They state that
voluntary actions (and decisions) should be seen as ‘the motor con-
sequences of processing and integrating large numbers of qualitatively
different types of input’ (ibid., p. 1702). For example, a soccer player
has to integrate many inputs and strategically respond to them during
a game. They propose that in order to study voluntary action and
decision making in a better way, what is needed is not a reduction of
the number of external inputs, but rather an increase of the number of
THERE IS NO FREE WON’T 21
for ‘simple’ decisions, which are not the result of complex integration,
to be free?
It seems that the general problem with these definitions is that the
researchers do not make a distinction between actions and decisions
that are the result of reasons, and movements that are merely caused.
Philosophers generally consider voluntary actions and decisions to be
special because we, as agents, play an active role in the outcome and
what we decide and do is not merely a product of causal ‘blind’
forces. We are able to take reasons into account, deliberate about
them, and because of that we can effectively respond to facts and
circumstances in a way that we find reasonable or justified. We want
to make a difference between hurting someone as a result of an acci-
dent, and hurting someone because of a reason, e.g. disliking the
person. The central point is that, in cases in which we act for reasons,
our actions can be rationalized from our own point of view (Schlosser,
2014, p. 250). However, the alternative definitions of what it means to
act voluntarily that neuroscientists in the field provide do not capture
this essential distinction between ‘blind’ causation and acting for
reasons. In order to do this, the relative contribution of internal and
external factors does not seem to explain this difference, as also
Schlosser (2014) argued in his paper, and it is unclear how enhancing
the complexity of the decision contributes to this. Thus, the proposal
would be to study actions that are done for reasons.
Furthermore, it is interesting that the understanding of free will is
not questioned in the neuroscientific literature in the field. In the
current conception in which conscious initiation is central, it is
assumed that for a decision to be free it has to be initiated by a
22 L. ASMA
conscious self that exists separate from the brain. However, they do
not argue for their position nor give an account of how this would
work, and/or how this would make free will and conscious action
control possible. Furthermore, their understanding of free will does
not seem to track folk intuitions about free will (see, for an overview
of conducted surveys, Mele, 2012). The burden of proof seems to be
on those that defend a supernatural or substance dualist account of
free will, which are in this case the neuroscientists that study it.
Thus, alternative definitions of voluntariness and freedom are
necessary for neuroscientific experiments to provide a valuable contri-
bution to the debate on free will and free won’t. In order to do so, they
should at least acknowledge that a central aspect of free will is
For personal use only -- not for reproduction
References
Brass, M. & Haggard, P. (2007) To do or not to do: The neural signature of self-
control, Journal of Neuroscience, 27 (34), pp. 9141–9145.
THERE IS NO FREE WON’T 23
Caruso, G.D. (2012) Free Will and Consciousness: A Determinist Account of the
Illusion of Free Will, Lanham, MD: Lexington Books.
Filevich, E., Kühn, S. & Haggard, P. (2012) Intention inhibition in human action:
The power of ‘no’, Neuroscience and Biobehavioral Reviews, 36, pp. 1107–
1118.
Filevich, E., Kühn, S. & Haggard, P. (2013) There is no free won’t: Antecendent
brain activity predicts choices to inhibit, PLoS ONE, 8 (2), e53053.
Fischborn, M. (2016) Libet-style experiments, neuroscience, and libertarian free
will, Philosophical Psychology, DOI: 10.1080/09515089.2016.1141399.
Haggard, P. (2008) Human volition: Towards a neuroscience of will, Nature
Reviews Neuroscience, 9, pp. 934–946.
Haggard, P. & Eimer, M. (1999) On the relation between brain potentials and the
awareness of self-generated movements, Experimental Brain Research, 126, pp.
128–133.
Haggard, P. & Libet, B. (2001) Conscious intention and brain activity, Journal of
For personal use only -- not for reproduction